On 10.Dec.2025, the Rights on the Net Coalition – of which Transparência Brasil is a member – published an open letter to congressmen and women calling urgently for the text of the report on Bill 2338/2023 (Artificial Intelligence Bill) to be made available by the rapporteur, federal deputy Aguinaldo Ribeiro, and for time to be allowed for analysis, so that civil society can look into any changes that may be made to the version of the text approved by the Federal Senate in 2024.

The communication was signed by 49 organizations, including TB:

Your Excellencies and Deputies,

Brazil cannot wait any longer to establish democratic rules for the development and operation of artificial intelligence in our country. The urgency of regulation is evident. Cases of irresponsible use of automated technologies have already caused serious damage in the country. In different Brazilian states, black people have been detained on the basis of facial recognition systems that had error rates of over 90 percent in false correspondence, as documented in surveys by state public defenders’ offices, the IDDD’s Prova sob Suspeita project and human rights organizations such as the Criminal Justice Network. This scenario disproportionately affects vulnerable citizens and demonstrates that the absence of minimum technical criteria creates immediate risks for the population.

The advance of digital fraud reinforces the seriousness of the situation. Cybersecurity reports indicate that scams involving AI-assisted social engineering, voice cloning and the creation of fake documents will grow by more than 300 percent between 2022 and 2024 in Brazil. The ease of use of these tools has already victimized the elderly, people with disabilities and low-income citizens, who have suffered financial losses and, in some cases, lost access to essential benefits after providing information induced by automated systems that simulated family members or public agents.

Within the criminal justice system, the adoption of artificial intelligence tools for risk analysis, predicting recidivism and supporting judicial decisions requires special attention. The use of these technologies without adequate regulation can perpetuate and amplify historical discriminatory biases, compromising the constitutional right to a full defense and due process of law. It is essential that any regulatory framework establishes specific rules for AI systems used in contexts of deprivation of liberty, criminal investigation and criminal execution.

In electoral periods, the absence of regulatory mechanisms aggravates the information environment. The ability to generate manipulated images, audio and video on a large scale puts the integrity of the public debate at risk. The country has already recorded the circulation of potentially falsified content that simulated speeches, statements and actions by political figures during the recent electoral calendar, which requires urgent regulatory responses to protect society and democracy.

Such regulation must be based on human rights, human dignity, equality and non-abusive discrimination. These principles should guide the creation of risk assessment procedures, damage mitigation, independent algorithmic auditing with the participation of civil society, and redress for people affected by automated decisions, including accessible mechanisms for contesting and human review of decisions, as well as transparency rules that allow synthetic content to be identified and the agents involved to be held accountable.

Copyright protection must also be ensured. Data mining for model training depends on protected works and generates significant economic value. It is essential that the framework recognizes and establishes mechanisms that guarantee fair remuneration to copyright holders, transparency about the use of works and the possibility of control by creators, avoiding misappropriation of national content.

We want an AI framework to promote and regulate technological development and innovation, based on universal ethical principles and standards, respecting and protecting human and fundamental rights, copyright, health, education, children and adolescents, the environment, citizenship, social justice and safeguarding democracy.

In April 2025, the Chamber of Deputies created the Special Commission on Artificial Intelligence in Brazil to debate Bill 2338/23, which provides for the development, promotion and ethical and responsible use of AI based on the centrality of the human ^person1.

Made up of 33 full deputies and an equal number of alternates, the Commission’s first meeting and installation took place on May 20, 2025, when it was set to present its Work Plan with the definition of a timetable and ^procedures2.

It is commendable that the Chamber of Deputies is guaranteeing a qualified debate on an issue of such urgency and which affects Brazilian society widely, seeking to include voices in the promotion of a genuine public debate, continuing the work of the Federal Senate. Despite this, the spaces for listening to civil society throughout this process have been extremely limited. A representative debate requires not just one or a few members of these segments, but a numerically equal, or at least balanced, composition of all the segments invited.

The first public hearing, for example, which took place on June 10, 2025, was attended by only one woman, one member of a civil society organization and no black, indigenous or other socially diverse people, precisely the groups most affected by the already documented risks of AI systems, as demonstrated by the cases of false identification by facial recognition that predominantly affect the black population. An unrepresentative debate will not allow the drafting of a regulation capable of meeting the different interests of Brazilian society. The under-participation of civil society, in particular, has the effect of compromising discussions by not adequately bringing in exactly the perspective of those who will be most affected by artificial intelligence, which are the people who use it.

We know that the deputies who make up the Commission have a deep commitment to and appreciation of Brazilian society, which is precisely why we ask for their attention to these points. In this way, we make ourselves available and reiterate our interest in collaborating towards the rapid conclusion of the work of the distinguished Special Commission, so that we can arrive at an AI regulatory framework for Brazil that protects fundamental rights and stimulates responsible innovation in the country.

¹ Chamber of Deputies. Hugo Motta creates committees to discuss PNE, income tax exemption and artificial intelligence. Agência Câmara de Notícias. Available at: https://www.camara.leg.br/noticias/1147592-hugo-motta-cria-comissoes-para-discutir-pne-isencao-do-imposto-de-renda-e-inteligencia-artificial/. Accessed May 19, 2025; MOTTA, Hugo. This ad is not the result of AI! Congresswoman @luisa_canziani (PSD-PR) will chair the Special Committee that will discuss the Artificial Intelligence project in the Chamber. The rapporteur will be my fellow deputy @depaguinaldo11 (PP-PB). Brasília, April 7, 2025. Disponível em: https://x.com/HugoMottaPB/status/1909390443755844034. Accessed on May 20, 2025.

² Chamber of Deputies. Chamber installs commission to debate proposal on the use of artificial intelligence in Brazil. Chamber News Agency. Available at: https://www.camara.leg.br/noticias/1159111-camara-instala-comissao-para-debater-proposta. Accessed on May 19, 2025.

Signatory organizations:

1. ABRA – Brazilian Association of Screenwriters
2. AMC-SP – São Paulo Film Editors Association
3. AnaMid – National Association of the Digital Market and Industry
4. AqualtuneLab
5. ARTICLE 19 Brazil and South America
6. Brazilian Association of Audiovisual Composers – Musimagem Brasil
7. Digital Journalism Association – Ajor
8. EIG Women’s Association – Evangelicals for Gender Equality
9. Women in Communication Association (AMC)
10. World Association of Community Radios – AMARC Brasil
11. CDDH Dom Tomás Balduíno
12. Center for Security and Citizenship Studies (CESeC)
13. Rights on the Net Coalition
14. Conectas Human Rights
15. Digital Collective – Association for the Democratization of Access to the Information Society
16. CTI – Renato Archer
17. data_labe
18. Data Privacy Brazil
19. Unveil
20. DiraCom – Right to Communication and Democracy
21. Dublar – Movimento Dublagem Viva
22. EDHIA/DPU
23. Geftas – Social Service and Communication
24. Research group Media, knowledge and the environment: views from the Amazon
25. Brazilian Institute of Social and Economic Analysis – IBase
26. Institute of Science Education Technology and Culture in Communities – ICETEC BRASIL
27. Institute for the Defense of Consumers (IDEC)
28. Institute for the Defense of the Right to Defense – IDDD
29. Internet and Society Reference Institute – IRIS
30. Instituto Soma Brasil
31. Sumaúma Institute
32. Telecom Institute
33. Intervozes – Brazil Social Communication Collective
34. IP.rec – Recife Institute for Research in Law and Technology
35. Communication Policies Laboratory of the University of Brasilia – Lapcom-UnB
36. Public Policy and Internet Laboratory – LAPIN
37. Laboratory of Work, Platformization and Health (LATRAPS-UEM)
38. Mangue Journalism
39. United Movement of Street Vendors (MUCA)
40. Portal IMPRENSA
41. Criminal Justice Network (RJC)
42. Reporters Without Borders – RSF
43. Satedsp
44. Union of Professional Journalists of Rio Grande do Sul (SindJoRS)
45. Sleeping Giants Brazil
46. Transparency Brazil
47. Brazilian Women’s Union São Gonçalo dos Campos Branch
48. Democratic Union of Digital Artists (UNIDAD)
49. WITNESS